A statistician's view of weather modification technology (A Review)

Abstract

The present paper is motivated by the enactment, in October 1976, of Public Law 94-490 which directs the U.S. Secretary of Commerce to formulate an appropriate national policy on weather modification. The current widespread drought contributes to the interest in the present state of this art. Here, then, the question of national policy becomes combined with the necessity of a more or less immediate decision on the use of public funds to pay for cloud seeding operations (as distinct from experiments) with the hope of increasing precipitation where it is most needed. The following presents a series of facts deemed relevant both to the immediate question of alleviating the drought and to the more general problem of long-range national policy. The specific questions considered and the present writer's answers are as follows. Question i. Is the present cloud seeding technology reliably confirmed as a means of alleviating drought? Answer: No. Question ii. Is there evidence that cloud seeding affects precipitation and, if so, what are the indicated effects? Answer: It appears established that cloud seeding does affect precipitation and does so over areas far in excess of the intended targets, occasionally up to distances on the order of hundreds of kilometers. In some cases the effects are large increases and in some others large decreases in precipitation. The several hypothetical mechanisms advanced to explain these effects and to predict them vary in their empirical support and convincingness. In particular, much of the existing literature, some of it stemming from official sources, is slanted and unreliable. Question iii. What are the means of advancing the development of a reliable weather modification technology? Answer: Establishment of at least two philosophically different interdisciplinary research groups, including statisticians versed in experimental work, perhaps members of the National Academy of Sciences, with a special mission to reevaluate the data of as many already performed cloud seeding experiments as possible, and continuation of properly planned experimentation. The suggested research groups should have unlimited access to the same data and have facilities for personal meetings to exchange ideas. They should be funded from sources other than those engaged in funding cloud seeding. Remark. The preliminary draft of this paper was sent to a number of knowledgeable persons with a request for comments. Some of the comments received caused certain changes in the original formulation. Some other comments indicated the desirability of response. Brief summaries of such comments, followed by my response, are interspersed in the paper. The beginnings and the ends of such insertions are marked by the symbol [CR]. The author is very grateful to all these scientists who read the original draft and summarized their reactions.